PY 81 as power rectifiers?

[Kalee20]

I'm including the update on the attached diagram, one erratum is the cathode capacitor is 470uF not 47uF as the diagram, its needed to stop + feedback.

That's very big for an AF cathode decoupler! Particularly as you seem to be experiencing HF rather than LF instability.

I will try your ideas on the coupling capacitors, the 470uF was at hand but it seems the trouble was because of the cathode resistors, I will try some different decoupling and possibly a grid stopper for the triode too, it may help. Also if I get the grid biased spot on it may help.

In this application, 470μF is quite reasonable. If it's to hand... use it. Modern 470μF capacitors are not expensive and not big, either, even if 'traditionally' smaller values were used.

There is a good reason here for going big, however. If you have insufficient decoupling, there will be feedback from the pentode to the TRIODE, because of the common cathode, and this feedback will be positive. You will get motorboating. Without the capacitor, the circuit forms a cathode-coupled multivibrator. Snip it out and try it!

Is there any reason a preset pot couldn't be used to set the bias empherically (sounds better than experimentally)

None whatever! Sometimes you can't beat a good twiddle with a screwdriver. I strongly encourage it - you will get to hear, and recognise, the sound of different types of distortion caused by under-biasing and over-biasing. But I'd suggest something like the sketch, so you don't need to use a preset pot which is rated for the full cathode current.